Automatic knitting machines use banks of large numbers of closely spaced latch needles to interlock threads in a series of connected loops to produce a knitted fabric. The latch needle is a long flat needle having, at one end, a small hook and a latch that swivels to open and close the hook. The hook ends of the latch needles are moved forwards and backwards towards and away from the threads being knitted into the fabric. As a latch needle is moved, its latch alternately opens and closes so that the hook catches a thread close to it, pulls it to create a loop of fabric, and then releases the thread to start the cycle over again and produce another loop of fabric.
Latch needles are arranged parallel to each other, in arrays of many hundreds to thousands of latch needles in modern knitting machines. The latch needles are placed into narrow latch needle slots that are machined into a planar surface, hereafter referred to as a "needle bed surface", of a large rectangular metal plate, hereafter referred to as a "needle bed". The latch needle slots hold the latch needles in position and confine their motion to linear displacements along the lengths of the latch needle slots. The latch needle slots are parallel to each other and equally spaced one from the other with spacing that varies depending upon the quality and type of fabric being produced. Spacing of two to three millimeters is typical, but spacing significantly less than and greater than two millimeters are also common.
The latch needle slots in a needle bed are sufficiently deep so that all or most of the body of a latch needle lies completely in the latch needle slot in which it is placed and below the needle bed surface into which the latch needle slots are machined. A small square fin that sticks out from one side of the shaft of the latch needle protrudes above the needle bed surface. The fins of all latch needles in a needle bed are accurately aligned in a single straight row perpendicular to the latch needle slots.
The latch needles are moved, hereafter referred to as "activated", back and forth in their respective latch needle slots in order to form loops in a fabric being knitted, by a shuttle that travels back and forth along the length of the needle bed surface parallel to the row of aligned latch needle fins. The shuttle has a flat planar surface facing and parallel to the needle bed surface that extends the full length of the shuttle along the direction of travel of the shuttle. The surface has a channel extending the full length of the shuttle along the direction of travel of the shuttle. The channel is open at both of its two ends, and both ends are aligned with the row of aligned fins. As the shuttle moves along the row of latch needle fins, the fins of the latch needles sequentially enter the channel at one end of the channel, travel along the channel length and exit the channel at the other end of the channel. For most of its length the channel is parallel to the row of aligned fins, i.e. the direction of travel of the shuttle, however towards its middle it has a bend. A latch needle is activated when its fin encounters the bend and moves along the direction of the bend. In moving along the direction of the bend, the fin and its latch needle are moved back and forth along the direction of the latch needle slot in which the latch needle is placed, i.e. perpendicular to the row of aligned fins.
The conventional method for moving latch needles in a knitting machine as described above has a number of drawbacks.
For one, the sequential activation of latch needles by a shuttle as the shuttle moves along a needle bed limits the production rates of fabrics. Production rates of fabric produced by knitting machines could be increased if latch needles were individually activated and different combinations of latch needles could be moved simultaneously. Some shuttles in fact have more than one channel in order to simultaneously activate more than one latch needle and increase production rate.
In addition, in the process of knitting a fabric, dust and dirt accumulate in the slots in which latch needles of a knitting machine move. As the dust and dirt accumulate, more force is required to move the latch needles. At some point, dust and dirt accumulate to such an extent that a latch needle jams in its slot. The shuttle is too massive and moves too quickly for it to be practical for the shuttle to be sensitive to, or respond to, changes in the force needed to move a particular latch needle. As the shuttle rushes along the needle bed and encounters a jammed latch needle it breaks the fin or some other part of the jammed latch needle. When this happens physical damage to the knitting machine is often considerably more extensive than the damage to the single latch needle that jammed and knitting machine down time as a result of the damage is prolonged.
In order to prevent damage to knitting machines from jammed latch needles it would be advantageous to have a system for moving latch needles in a knitting machine that activates latch needles individually and is responsive to changes in the forces required to move individual latch needles.
Prior art direct needle drive systems exist that provide for individual activation of latch needles in a knitting machine. These systems, hereafter referred to as "DND" systems, generally provide an actuator for each latch needle and a system for monitoring the position of each latch needle. However, the prior art systems have not been completely satisfactory. The dimensions of actuators used in the prior art systems are large compared to the spacing between latch needles. Complicated spatial configurations are therefore required to pack large numbers of the actuators in a convenient volume of space near to the latch needles in order to couple the actuators to the latch needles.
Additionally, the response times of prior art DND systems are slow. This is the result of slow response times of actuators and of latch needle position monitoring systems used in these systems. The advantages in production rate and decreased knitting machine down time that should be provided by prior art DND systems are at least partly neutralized by the slow response times of these systems.